
The validity of our Monte Carlo simulation procedure ({\it the integral
method}) had been verified by the corresponding analytical procedure of which
is quite independent of our method methodologically. Also, the results obtained
by our procedure are compared with those obtained by the different Monte Carlo
simulation procedure ({\it the differential method}) which have been
exclusively utilized by the different authors and the agreement between them
are found to be well. By utilizing our Monte Carlo procedures, the validity of
which is guaranteed in two different procedures, we investigate not only the
fluctuation of high energy muons themselves but also fluctuation of the various
quantities related to the energy losses by the muons, which are difficult to
obtain by {\it the differential method}. Namely, we obtain fluctuation on
energy losses of the muons, fluctuation on Cherenkov lights due to the
accompanied cascade showers initiated by the muon and the correlations between
them . Finally, we obtain the transition curves for Cherenkov lights in KM3
detector, taking into account of all possible fluctuations in the stochastic
processes and point out the difficulty of the reliable estimation of the energy
of the muons which are resultants of muon neutrino events in the KM3 detectors.

Adopting neutrino oscillation parameters obtained by SuperKamiokande, a
numerical computer experiment for neutrino events occurring outside the
detector, is carried out in the same SK live days, 1645.9 live days,
constructing the virtual SuperKamiokande detector in the computer. The
numerical results by the computer experiment could be directly compared with
the real SK experimental data. The comparison shows that it is difficult to
obtain convincing conclusion on the existence of the neutrino oscillation with
specified neutrino oscillation parameters claimed by SK through analysis for
neutrino events occurring outside the detector.

The energy spectra of hadron cascade showers produced by the cosmic ray muons
travelling through water as well as the muon energy spectra underwater at the
depth up to 4 km are calculated with two models of muon inelastic scattering on
nuclei, the recent hybrid model (twocomponent, 2C) and the wellknown
generalized ectormesondominance model for the comparison. The 2C model
involves photonuclear interactions at low and moderate virtualities as well as
the hard scattering including the weak neutral current processes. For the muon
scattering off nuclei substantial uclear effects, shadowing, nuclear binding
and Fermi motion of nucleons are taken into account. It is shown that deep
nderwater muon energy spectrum calculated with the 2C model are noticeably
distorted at energies above 100 TeV as compared to that obtained with the GVMD
model.

The SuperKamiokande group assert that they have found an oscillatory
signature in atmospheric neutrinos through the analysis of \textit{Fully
Contained Events} and \textit{Partially Contained Events}. We have performed an
$L/E$ (length/energy) analysis of \textit{Upward ThroughGoing Muon Events} and
\textit{Stopping Muon Events} in a numerical computer simulations both with and
without neutrino oscillations but were unable to find an oscillatory signature.
We give likely explanations for the absence of the oscillatory signature in our
simulations and its apparent presence in the SuperKamiokande data.

Differential and integral cross sections of the muon are calculated in the
materials: water, standard rock, iron and lead with and without the LPM effect.
The corresponding cross sections are also calculated with dielectric supression
effect (TerMikaelian effect), in addition to the LPM effect. To demonstrate
the importance of the LPM effect in the bremsstrahlung process, depth intensity
relation of muon, energy spectrum, range distribution and survival probability
are calculated taking into account bremsstrahlung spectrum.